Thank you very much. It's such a great honor to be here to deliver one of the 2024 Nobel lectures, in large part because it means we've been awarded the Nobel Prize, which is a fantastic, fantastic honor. But rather than talk more about my joy about this, let me delve in and talk about institutions, technology, and prosperity. I'm going to take this opportunity to tell you a little bit about how I got interested in those questions, how Simon, Jim, and I approached them, but also provide a framework that I think is useful for thinking not just about history but about some questions of import today. What through economics can be seen from this picture? Even as a teenager, I became aware that there were huge differences across countries. So, for example, in this map, if you look at the countries that are shown in dark blue, which are the 12 richest countries, or 12 richest countries that make up about the top 10% of the distribution by GDP per capita around the world, they are about 63 times as rich as the countries shown in the lightest color. That's a staggering gap in a hyperconnected, globalized world, and it constitutes, to me, one of the most puzzling but one of the most exciting questions of social science.
Even as a teenager, I had reason to link this to politics because I was growing up in Turkey when or shortly after the country had undergone a coup, and freedoms were suppressed. Soldiers were everywhere, even at the gates of our school, and I came to wonder whether political issues were related to the economic problems that the country was experiencing at the time, such as unemployment, inflation, stagnation, and whether they were related to the issues that I was also observing around the world in my little capacity that I had as a teenager. So today we would say that those issues are related to institutions, broadly speaking: property rights, laws, regulations, political arrangements that influence the distribution of political power—who has political power, how that can be exercised—and the distribution of economic resources, which shape incentives and opportunities.
You can see on the right a huge variation in institutions, however you measure them. Here I'm using the World Bank's index of rule of law, equality before the law, and you can see some countries cluster at the bottom—I'm showing them in purple—and some at the very end, showing in green. We call those different institutions extractive and inclusive, just to have a label. Extractive institutions centrally are those that concentrate economic and political power in the hands of a small segment of society without putting much constraints on it, and that enable extraction of resources, tilting of the playing field in favor of those at the expense of others. Inclusive institutions, sort of the polar opposite, they distribute economic and political power and opportunities more broadly. You can see a strong correlation with GDP per capita, but correlation, of course, is not causation. In the parlance of economics, we would say institutions are endogenous. It may well be that, as John said, rich countries demand or can afford better institutions, whatever dimension of it you focus on, or it could be that there are some other omitted factors that are responsible, that influence both prosperity and various dimensions of institutions.
These ideas go back a long time. Montesquieu was one of the people who eloquently, although idiosyncratically, identified one version of it, which we later refer to as a geography hypothesis. He wrote, 'The heat of the climate may be so excessive as to deprive the body of all vigor and strength, and continued slavery is more supportable than the force and vigor of mind necessary for human conduct.' From this, Montesquieu concluded that his theory of separation of powers and republican forms, constitutional forms, was applicable to people who lived in cold climates, but not to those in hot climates, who deserved and would end up with despotism. So perhaps that's what we're seeing. How can we cut through this quagmire? Well, you need institutions, and you need context, and you need theory. So European colonial history seemed like an ideal lab for us, not just because it's one of the most formative events of the last millennium, but because, as the figure I showed already indicates, and I'll go back to it, the range of institutions you see in the former colonies spans the whole spectrum. So here, the countries that I picked at the bottom and the top—and they're not the only ones, by the way—are all former colonies. So you see this variation in the colonial sample as you see in the whole world.
But also we need a much more specific theory. In fact, a two-layered theory: first, about how institutions evolve and influence economic outcomes, and secondly, about why institutions vary. Why do you see that whole range of differences in the colonial world in terms of various dimensions of institutions? And that's the heart of what, in a more fancy way, you know, econometrician, statistician, economist would call an instrumental variable strategy, meaning you identify historically a source of variation that you can then exploit to explore the causal effect of institutions. So let me talk a little bit about that theory in a schematic form. Here is one representation: political institutions and the distribution of resources are what I call state variables. They come from the past, not in an unchanged form, but they come from the past, and they influence the de jure political power and the de facto political power. The de jure political power allocated by political institutions is, you know, who gets to vote, what are the rights, what are the representation, what are the constraints, and so on. But reality is that a lot of politics is not just about what is written on parchment, but it's about the reality of politics: how you form coalitions, how you influence others, how you use military resources, threats, and so on. Unfortunately, a very real part of institutions in reality, and it's the sum total of the de jure power and de facto power coming from who is rich, who is poor, who is enabled, who is empowered, that determines collective decisions. And the collective decisions that we most care about are about institutions. Economic institutions are shaped by this struggle, and the red arrow there is what we're interested in: the impact of economic institutions on economic performance, something like aggregates like GDP per capita growth, but also who gets the benefits, the distribution of resources in the future, and also the sum total of the de jure and de facto political power also shaped the evolution of institutions. Institutions are persistent, but they also change. There is nothing historically static about political institutions.
When we come to thinking about why was it that big variation existed between countries that were colonized by Europeans, we needed sort of a perspective too. And the way that Simon and Jim and I thought about this is summarizing the schema, but here let me start from the end, Rod, which is the link from current institutions to current performance. That's what we're interested in estimating. Link C, from early institutions to current institutions, is like how do colonial institutions, arrangements, power, and so on affect the organization of these societies today. And link B was the heart of our theory, or one part of the heart of our theory, which is the codetermination of settlements and early institutions. Europeans settled in some places and set up indirect rule or just tried to rule with a few people there over a large indigenous population. But of course, settlements themselves are determined endogenously. People were more likely to settle in places that offered greater riches.
So the approach that we developed was to think of, to look for something that would in turn be more exogenous to these dynamics, but still influence these settlements, and that was, of course, the local disease environment. Europeans had almost no immunity to yellow fever, malaria, many gastrointestinal diseases from the semitropical areas. So in some places where such diseases were absent, living conditions were more comfortable, healthier than even in Europe. But in some places—Africa, Central America, parts of Asia—Europeans faced staggeringly high mortality rates. So the chain of causation, therefore, is from these local conditions to settlements, co-determining institutions, which in a way that I'm going to explain in a little bit, once I provide the framework, then those persisting to some extent to current institutions, and then enabling us to estimate that link D. But the key here is what econometricians or we would call exclusion restriction, meaning that this potential settler mortality that Europeans faced at the time was not directly affecting current performance, but it's affecting through these channels. So that's the theory plus data, and in our case that's plausible because the lack of knowledge about germ theory of disease, about these infectious diseases and those immunities, those are long in the past, hundreds of years ago. So now the main channel is through these things.
So of course, a hard part of any empirical project is the data. So once we collected the data, we cleaned it and developed some degree of trust in it. So we looked at a figure like this, which you would call the first stage of this instrumental variable strategy, meaning the relationship between the endogenous variable that we're interested in. Again, I'm using the rule of law; doesn't matter much which index of institutions you use, actually, against the potential settler mortality in the colonial era. Potential here I'm emphasizing because in some places Europeans didn't actually settle, so these are numbers from military expeditions and such, so on. And you see a very strong relationship here, which when we first discovered it, this was the eureka moment for us. Oh my God, there is something here that's actually pretty big. So there is some degree, some quite high degree of projection of these current-day institutions to long-ago settler mortality rates. But our eureka turned into euphoria when we looked at this, which we call the reduced form, meaning this is the relationship between this source of variation, potential settler mortality, and the outcome of interest, GDP per capita today, as a result of these cumulative development processes. You see a very strong relationship; almost 50% of the variation is still projecting on these historical variables. And the exclusion restriction that I mentioned essentially says this is all working through those channels that I highlighted, not through some direct effect of this settler mortality. So put these two together, and this is the bottom line. The horizontal axis is again the rule of law index from the World Bank, and on the GDP, on the vertical axis, I have GDP per capita. But this is not the correlation that I showed you earlier; it is our estimate of the causal effect of institutions superimposed over this cloud, and it's the causal effect that we estimate through these assumptions and using the combination of theory and data.
From the beginning, my interest in economics was in technology: how it's shaped, how it influences growth, including inequality, who benefits from growth. And Simon, Jim, and I realized, of course, one of the major channels by which institutional arrangements mattered was by encouraging or discouraging technology adoption, how efficiently the economy was organized. And I talk about some of those, but I'm showing two summary measures again as the IV estimate using the same procedure that I described before. Again, these are not correlations, but these estimates superimposed on the data. On the left, I'm using a favorite measure of economists: total factor productivity. Look at those GDP per capita differ, but take out what is due to machinery, investments in physical capital, investments in human capital. What's left, the residual, that again projects onto the GDP, onto institutional variables, instrumented exploiting this exogenous source of variation. And on the right, a different measure that looks at the technology composition of exports: what is the fraction of exports of a country that are in products that are classified as advanced technology products by the Census Bureau. Again, you see a high projection. So whatever is working is working by co-determining technology, efficiency of production, and GDP per capita.
So to think about these, what I'm going to do is now I'm going to present a framework, and the point of this framework is to reinterpret our work, including the colonial context, but also help us think more about institutional change and institutional persistence, but also shed light on some questions of relevance today. For example, how we are going to adapt, benefit, or not benefit from a slew of new technologies we're experiencing in the context of AI. The heart, the centerpiece of this is what I'm going to call the utility-technology possibilities frontier, UTF, and this is going to describe how various social choices determine the distribution of utility, income between different groups. And using this framework, I can talk about factors that shift the frontier more or less efficiently, for example, technology, factors that move us along the frontier, and the role of technologies, especially in shifting the frontier. But also, this framework is going to be useful for delineating the causes of institutional persistence and institutional change. So here is what a UTF would look like in an extremely simple world with just two groups, which I've labeled poor and rich, just for concreteness. You can think of the rich as the colonialists, landowners, capitalists, or poor as indigenous population, peasants, or workers, depending on context. And think of the rich as being initially politically more powerful. The utility of the rich is on the vertical axis, that of the poor on the horizontal axis. So you see the frontier, the downward-sloping part of it, which is saying, well, over along this frontier, you can either provide more utility, more income to the poor, or more to the rich.
So let's get into the details a little bit and think about the moves and shifts and try to illustrate what this actually means. And here I have shown first along the blue frontier a move. There are two points there, B and A, and this is the trade-off that I was already hinting at. Point B has higher payoff for the poor and lower payoff for the rich, and a higher payoff for the rich and relatively lower payoff for the poor, as you can see from the projections on the horizontal and the vertical axis. What does this mean? So, for example, point B might be unions are strong, they negotiate high wages, and that leaves less money for the capital owners and perhaps the managers. But then moves along the frontiers are about the balance of power. If your unions weaken, that means less bargaining, lower wages, and profits increase, wages fall, and that is the move along the frontier from point B to A. But in the colonial period, in our work, for example, in Why Nations Fail with Jim, we emphasized a lot blocking of technology, use of inefficient technology. That's about shifts of the frontier, which has quite major consequences for, you know, the capacity of the economy to generate economic growth.
So the key is the shifts of the institutions as well as the moves along it. And so institutions shifting the frontier. Why would that be so? Hold up. That's one potential label you can give to the fact that the politically powerful groups cannot refrain from grabbing, extracting the income of others in the population, which will then discourage investment, innovation, economic participation by these other groups. This is very much about institutions. Why? Because when they do such extraction, that's a violation of the property rights of others or taking away opportunities, and it is the political institutions in part that determine whether they can do so or not. Constraints on the exercise of power would generate this sort of hold up and take resources from the others. In a lot of our work, we also thought about state capacity, the ability of the state to achieve its objectives, including enforcing laws and regulations. Extractive institutions also heighten conflict, and in doing so, they often erode or prevent the building up of state capacity. They sometimes lead to the collapse of state capacity. That's another channel via which you can get the kind of shifts from the blue to the red frontier, from the better to the worse frontier.
Some of my work recently is thinking about experimentation and the building of collective knowledge. Institutions are particularly important in this context, but I'm not going to have much time to talk about it. But I just paid lip service to it. But I'm going to spend a little bit more time on two channels that Jim and I emphasized in our work, which we called economic losers and political losers, which explain how this frontier is useful for thinking about these issues. Economic losers is about the distribution of and efficiency not being separable, they're being co-determined. So imagine we start at point C with the rich elite being powerful and getting a pretty good deal in terms of economics. For example, they control some monopolies. That's inefficient. So you can dismantle the monopolies and the distortions reduce. There's more competition, so you shift from the red to the blue frontier, or you introduce a new technology. But when you go to the blue frontier, the incomes of those who were previously benefiting from the monopoly, some of the politically most powerful people, will be eroded. New technology may require new talents, new comparative advantages. It won't be the existing elite that benefit from them. So that might take us to point B, and we don't have the fiscal tools such as lump-sum taxes to efficiently distribute these resources back. So if that's the case, from the point of view of the rich elite, you have a move from point C to point B, a good shift, but also losing of your economic rents and economic sources of benefits. And point C compared to point B, and look at the whole vertical projections of these, and you'll see actually the society as a whole gains, but the elite loses out. That's the economic losers mechanism. So the fact that they are going to lose their economic privilege makes them reticent to embrace institutional change or new technologies, or it might even force them sometimes to choose worse institutions when better ones are available.
But the problem is not just about fiscal tools, and in fact, the political losers, what we call the political losers mechanism, is, we believe, is more fundamental. So here again, we start at C, the same monopolies, the same bad technology, and better alternatives are available. We shift to the blue frontier, but now the instruments are available, so you don't have to go to point B. You can go to point A, the best of all worlds. Both the rich elite and the citizens benefit. But here is the catch: change, economic change begets political change. This is what Jim and I called in Why Nations Fail, political creative destruction. Some discontent is created that creates political instability, or even more importantly, the foundation of the political power of the elite may be eroded by economic change. So as a result, you expect the flow, you lose political power, and once you lose political power, the same fiscal tools are now turned against you. So you start traveling from point A to B. So the choice may be between C all the time versus A today versus B tomorrow. And if you expect to lose a lot from your loss of political control, you will be in this political losers situation. You oppose political change because of fear of loss of political control. So we think it's more fundamental, and in fact, it's very much interwoven with the theory of economic change, the theory of political change that John mentioned, how democracy emerges. Sometimes peacefully, it's about the ability of different people to make concessions and arrange the gains from economic interaction.
So again, start at point C, but remember that the elite controls institutional power, de jure power, but sometimes the more numerous rest of society may get de facto power. They may solve their collective action problem. They may pose a revolutionary threat, and when that happens, the situation is dire. So you have to make some concessions to stave off that revolution as an elite. So that concession may be move from point C to point B. The elite still controls the de jure power, but that's now a disadvantage because they're making concessions. But everybody understands that those concessions are temporary, that the revolutionary threat is here today, gone tomorrow. And when that threat goes, you will draw back those concessions. So go back from point B to point C. So what do you do with this? So one way of dealing with the situation is to actually change the distribution of de jure power. Institutional change is a commitment to actually making these concessions more durable. So that was the heart of our explanation, and this is what, for example, the British politicians in the context of the First Reform Act, the first major piece of democratic legislation in the 19th century, when Lord Grey introduced to the parliament, the principle of my reform is to prevent the necessity of revolution. These frameworks frame the way that we thought about the colonial era as well as others.
So let me talk a little bit about the colonial era. Understanding extractive institutions is quite straightforward. Extractive colonial institutions were characterized by a huge imbalance of power, militarily and otherwise, in the hands of the Europeans, but limited fiscal tools for extractive extraction that would be more efficient. So they set up hugely inefficient, hugely repressive monopolies, extractions, coercion systems for labor extraction, and a small elite ruling over a large land mass and large population. Political control was paramount in their mind, so political losers mechanism was central. So it's no surprise to us from this framework that you ended up with highly extractive institutions in the colonial era in all of these places, which the Europeans tried to control, especially in the places they didn't settle. But why different in the settler colonies? Not because Europeans came up with better models of democracy and implemented them, and not because colonial authorities had different designs. In fact, we documented that they tried to do often the same things as they did in Spanish America, in North America, but it didn't succeed. It didn't succeed because the balance of power was not as skewed. First of all, the settlers could move to other places, the open frontier in this often relatively empty land compared to the places like the Inca, Aztec, or India. But also, Europeans, colonial authorities were less willing to exact the same violent repression on Europeans as on the native population. So that more balance also meant that often they had to make concessions to the lower strata, which were these indentured servants and so on. You see that in the Jamestown Colony in Virginia, Pennsylvania, in Australia. First, more inclusive, a little more inclusive economic institutions, more inclusive political institutions, some legislatures and so on, and that started kick-starting a very different process.
But if I want to talk more about the colonial era and other episodes, another concept that is useful, and I'll introduce that briefly, this is what we called critical junctures: periods of disruption because of demographic change, new economic opportunities, new technologies that create room for both institutional and economic change. And the way that I conceptualize that is something that leads to major divergences from small differences. And so again, look at the frontier with the red. I'm showing the frontier under extractive institutions. The blue is the shift, the better frontier when you have more inclusive institutions. But in regular times, but small change becomes much amplified during these critical junctures. So that's why I am drawing in green what the same institutions would achieve during this period of critical juncture. Why? There are new opportunities, so changes now enable much more beneficial variations in terms of economics, in terms of who runs the firms and who invests and who innovates and so on. But critical junctures are also at times of institutional divergence. If you try to increase repression in order to deal with these new opportunities, new instabilities, that's a shift inwards. That's the black frontier, path dependent, but also very divergent changes, how we try to conceptualize it. And you see once you do this, new commonalities, for example, between colonialism and the spread of industrialization. In colonialism, it was these small differences that led to huge variations in terms of institutions. Before Europeans arrived, there were differences in terms of institutions, in terms of some civilizations that were pretty elaborate, and there were some income differences in the colonial world. But these small differences became amplified. Why? Some relatively inconsequential things like the local diseases that didn't matter so much for the indigenous population suddenly had a much different meaning in terms of institutional trajectories when Europeans, with their huge power imbalance but their lack of immunities, arrived. Industrialization within the colonial world can be understood in the same way.
But let me actually talk about industrialization within Europe, within Asia, within the colonial world. You see these divergences during industrialization: some countries taking advantage of industrial opportunities, railways, new factories; some others resisting them. Why? Well, again, small differences becoming amplified via institutional trajectories diverging via different reactions to technologies and how that shapes the path of technology. So France and the Habsburg Empire weren't that different in the 18th century, but differences in the political dynamics meant that the political losers mechanism was much more of a concern for the Habsburg dynasty. Or as a leading lawmaker put it, 'We do not desire at all that the great masses shall become well off and independent. How could we rule over them otherwise?' So that meant that these new opportunities were met with greater repression and refusal from the Habsburgs, and not so from France or some other countries. Institutional divergence, economic divergence, and institutional responses, for example, during the Progressive Era, the balance of power shifting back and forth between monopolies and not.
But what I want to do is to end by applying this framework to our current questions, and there are so many of them, and there are questions that we could think about and reason perhaps using this framework, perhaps with others. But one that I think is at the back of the minds of many people is about digital technologies, new inequalities, and the age that may be shaped as the age of AI. So again, let me start thinking about this using the red frontier where we were, and imagine we were at a point like B0, where there are relatively high wages, shared period of growth throughout the three decades that followed World War II. And for a variety of reasons, balance of power starts shifting, deindustrialization, unions becoming weaker, and that's a shift from B0 towards B1. Now think of B1 as favoring the capital owners and the professional workers, the highly educated who become the managers, at the expense of manual workers. But the key is again technology and shifts of the frontier. So, for example, in my work with Pascual Restrepo, we have emphasized how you cannot understand the current period without thinking of automation and how technologies have changed. And automation is particularly important because it shifts tasks away from labor, especially manual, low-education workers, towards capital. So it's really shifting and tilting the frontier, and that's what I'm showing here with the blue frontier. It's tilting in favor of capital owners and professional workers and against workers. That's one conceptualization of automation. So now the situation is much worse because you're not just moving along the frontier, but the whole shift is against the manual workers.
But the age of AI is intensifying these trends. But what I have emphasized in my work, and I firmly believe, and for example what I have written with Simon Johnson, with Pascual Restrepo, with David Autor, is that AI is a versatile technology. You can use them in many different ways. You can use AI to provide better information to make workers, manual workers, much more productive. That's what we call pro-worker AI. And if you do that, you can go back to something like the black frontier, which provides a chance for building shared prosperity. Wages, utilities, incomes of different segments of population can be high. But the fear is that it can also lead to the green frontier, even more distorted, but also not so favorable. This is what we call excessive automation. This is very bad for workers, but not so good for the capitalists and professional workers either, because when you're doing the excessive automation, you are not benefiting because it's excessive. You're actually losing some of the productivity. It is a good way of controlling labor. It's not a great thing for labor, but it's not so beneficial either. So there are choices that are political as well as technical about how we develop AI and how we use it, which is going to be critical not just for the distribution of resources, but the critical juncture perspective says about the evolution of institutions. If you create different types of technologies and different kinds of distributions of gains, that also has implications about who's going to be empowered and how democracy will function and so on.
So in conclusion, I am really excited to be here for many reasons, but one of them is to be able to share what I'm really excited about, which is about political economy and of institutions and technology. And I think the framework as well as the summary of the work that I provided was meant to convey not just that enthusiasm, but that once you start looking at the world through these lenses, you see new commonalities, and there are lessons that we can draw from very disparate episodes such as the spread of industry, colonial period, for the current age. And I think not just these commonalities but others will be the heart of the political economy of technology and our understanding of the future of institutional divergences and institutions and technology. And I think many people in this room and many people around the world who are really excited about these issues, I think, are going to make amazing contributions here. And if we can continue to be part of this journey, Simon, Jim, and I would be amazingly happy. Thank you.
